Skip to content

Commit

Permalink
Refactored to separate corrected and raw wheel data completely in plot.
Browse files Browse the repository at this point in the history
  • Loading branch information
@adamshapiro0
adamshapiro0 committed Nov 21, 2023
1 parent 56a2f8c commit d294d6e
Showing 1 changed file with 40 additions and 47 deletions.
87 changes: 40 additions & 47 deletions python/fusion_engine_client/analysis/analyzer.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -1358,12 +1358,12 @@ def _plot_wheel_ticks_or_speeds(self, source, type):
figure_title = 'Measurements: %s Encoder Ticks' % source.title()

if source == 'wheel':
measurement_type = self._auto_detect_message_type([RawWheelTickOutput, WheelTickInput])
raw_measurement_type = self._auto_detect_message_type([RawWheelTickOutput, WheelTickInput])
else:
measurement_type = self._auto_detect_message_type([RawVehicleTickOutput, VehicleTickInput])
raw_measurement_type = self._auto_detect_message_type([RawVehicleTickOutput, VehicleTickInput])

# Wheel ticks are raw (uncorrected) by definition.
raw_measurement_type = measurement_type
# Wheel ticks are raw (uncorrected) by definition. There are no corrected wheel ticks.
measurement_type = None
else:
filename = '%s_speed' % source
figure_title = 'Measurements: %s Speed' % source.title()
Expand All @@ -1372,26 +1372,25 @@ def _plot_wheel_ticks_or_speeds(self, source, type):
measurement_type = self._auto_detect_message_type([WheelSpeedOutput, DeprecatedWheelSpeedMeasurement])
else:
measurement_type = self._auto_detect_message_type([VehicleSpeedOutput,
DeprecatedVehicleSpeedMeasurement])
DeprecatedVehicleSpeedMeasurement])

if source == 'wheel':
raw_measurement_type = self._auto_detect_message_type([RawWheelSpeedOutput, WheelSpeedInput])
else:
raw_measurement_type = self._auto_detect_message_type([RawVehicleSpeedOutput, VehicleSpeedInput])

if measurement_type is None:
measurement_type = raw_measurement_type

if measurement_type is None:
if measurement_type is None and raw_measurement_type is None:
self.logger.info('No %s %s data available. Skipping plot.' % (source, type))
return

any_measurement_type = measurement_type if measurement_type is not None else raw_measurement_type

# If the measurement data is very high rate, this plot may be very slow to generate for a multi-hour log.
if self.long_log_detected and self.truncate_data:
params = copy.deepcopy(self.params)
params['max_messages'] = 2
result = self.reader.read(message_types=measurement_type, remove_nan_times=False, **params)
data = result[measurement_type.MESSAGE_TYPE]
result = self.reader.read(message_types=any_measurement_type, remove_nan_times=False, **params)
data = result[any_measurement_type.MESSAGE_TYPE]
if len(data.measurement_time) == 2:
dt_sec = data.measurement_time[1] - data.measurement_time[0]
data_rate_hz = round(1.0 / dt_sec)
Expand Down Expand Up @@ -1434,12 +1433,12 @@ def _extract_data(measurement_type):
titles = ['%s Speed (Unsigned)' % source.title(), 'Gear/Direction']
titles.append('Measurement Interval')

if data is not None:
if data is None:
titles[0] += f'<br>Messages: {raw_measurement_type.__name__}'
elif raw_data is None:
titles[0] += f'<br>Messages: {measurement_type.__name__}'
if raw_data is not None and measurement_type != raw_measurement_type:
titles[0] += f', {raw_measurement_type.__name__}'
else:
titles[0] += f'<br>Messages: {raw_measurement_type.__name__}'
titles[0] += f'<br>Messages: {measurement_type.__name__}, {raw_measurement_type.__name__}'

figure = make_subplots(rows=len(titles), cols=1, print_grid=False, shared_xaxes=True, subplot_titles=titles)

Expand All @@ -1463,47 +1462,42 @@ def _extract_data(measurement_type):
#
# All output messages from the device should contain P1 time. We should only ever use a non-P1 time source when
# plotting logged input messages (uncommon).
def _get_time_source(measurement_type, data):
if measurement_type is None or data is None:
def _get_time_source(meas_type, data):
if meas_type is None or data is None:
return None
# If this data does not have P1 time, use its incoming native time source (system time of reception, etc.).
elif np.all(np.isnan(data.p1_time)):
# Check that the time source never changed. Warn if it did.
if np.any(np.diff(data.measurement_time_source) != 0):
self.logger.warning('Detected multiple time source types in %s data.' % measurement_type.__name__)
self.logger.warning('Detected multiple time source types in %s data.' % meas_type.__name__)

result = SystemTimeSource(data.measurement_time_source[0])
self.logger.warning('%s data does not have P1 time available. Plotting in %s time.' %
(measurement_type.__name__, self._time_source_to_display_name(result)))
(meas_type.__name__, self._time_source_to_display_name(result)))
# P1 time available - use that.
else:
result = SystemTimeSource.P1_TIME
return result

time_source = _get_time_source(measurement_type, data)

if raw_measurement_type is not None and measurement_type != raw_measurement_type:
raw_time_source = _get_time_source(raw_measurement_type, raw_data)
if time_source is None:
time_source = raw_time_source
elif time_source != raw_time_source:
self.logger.warning('Different time sources detected for %s and %s. Data may not align.' %
(measurement_type.__name__, raw_measurement_type.__name__))
else:
raw_time_source = time_source

same_time_source = True
if measurement_type == raw_measurement_type:
common_time_source = time_source
elif time_source is None:
if raw_measurement_type is None:
corrected_time_source = _get_time_source(measurement_type, data)
raw_time_source = None
common_time_source = corrected_time_source
elif measurement_type is None:
corrected_time_source = None
raw_time_source = _get_time_source(raw_measurement_type, data)
common_time_source = raw_time_source
elif raw_time_source is None:
common_time_source = time_source
else:
self.logger.warning('Both raw and corrected %s data present, but timestamped with different '
'sources. Plotted data may not align in time.' % source)
same_time_source = False
common_time_source = None
corrected_time_source = _get_time_source(measurement_type, data)
raw_time_source = _get_time_source(raw_measurement_type, raw_data)
if corrected_time_source == raw_time_source:
common_time_source = corrected_time_source
else:
common_time_source = corrected_time_source
same_time_source = False
self.logger.warning('Both raw and corrected %s data present, but timestamped with different '
'sources. Plotted data may not align in time.' % source)

if same_time_source:
time_name = self._time_source_to_display_name(common_time_source)
Expand All @@ -1513,16 +1507,16 @@ def _get_time_source(measurement_type, data):
for i in range(len(titles)):
figure['layout']['xaxis%d' % (i + 1)].update(title=time_label, showticklabels=True)
else:
time_name = self._time_source_to_display_name(time_source)
corrected_time_name = self._time_source_to_display_name(corrected_time_source)
raw_time_name = self._time_source_to_display_name(raw_time_source)
figure['layout']['annotations'][0]['text'] += '<br>Time Source: %s (Raw), %s (Corrected)' % \
(raw_time_name, time_name)
(raw_time_name, corrected_time_name)

time_label = f'{time_name}/{raw_time_name} Time (sec)'
time_label = f'{corrected_time_name}/{raw_time_name} Time (sec)'
for i in range(len(titles)):
figure['layout']['xaxis%d' % (i + 1)].update(title=time_label, showticklabels=True)

p1_time_present = (time_source == SystemTimeSource.P1_TIME or
p1_time_present = (corrected_time_source == SystemTimeSource.P1_TIME or
raw_time_source == SystemTimeSource.P1_TIME)

# If plotting speed data, try to plot the navigation engine's speed estimate for reference.
Expand Down Expand Up @@ -1687,9 +1681,8 @@ def _plot_vehicle_data(data, time_source, is_raw=False, show_gear=False, style=N
# Plot the data. If we have both corrected (e.g., WheelSpeedOutput) and uncorrected (e.g., RawWheelSpeedOutput)
# messages are present in the log, plot them both for comparison.
_plot_func = _plot_wheel_data if source == 'wheel' else _plot_vehicle_data
_plot_func(data, time_source, is_raw=measurement_type == raw_measurement_type, show_gear=True)
if measurement_type != raw_measurement_type:
_plot_func(raw_data, raw_time_source, is_raw=True, show_gear=False)
_plot_func(data, corrected_time_source, is_raw=False, show_gear=True)
_plot_func(raw_data, raw_time_source, is_raw=True, show_gear=False)

self._add_figure(name=filename, figure=figure, title=figure_title)

Expand Down

0 comments on commit d294d6e

Please sign in to comment.